Skip to main content

Table 1 Comparison against different methods on SCOP 1.53 superfamily benchmark

From: A discriminative method for protein remote homology detection and fold recognition combining Top-n-grams and latent semantic analysis

Average ROC and ROC50 scores
Methods ROC ROC50 Source
SVM-Top-n-gram    
n = 1 0.9069 0.696  
n = 2 0.9230 0.713  
n = 3 0.9073 0.653  
SVM-Top-n-gram-combine 0.9329 0.763  
SVM-Bprofile(Ph = 0.13) 0.9032 0.681 [28]
SVM-Ngram 0.7914 0.584 [32]
SVM-Pattern 0.8354 0.589 [32]
SVM-Motif 0.8136 0.616 [32]
SVM-Top-n-gram-combine-LSA 0.9390 0.767  
SVM-Bprofile-LSA(Ph = 0.13) 0.9210 0.698 [28]
SVM-Ngram-LSA 0.8595 0.628 [32]
SVM-Pattern-LSA 0.8789 0.626 [32]
SVM-Motif-LSA 0.8592 0.628 [32]
PSI-BLAST 0.6754 0.330 [32]
SVM-Pairwise 0.8960 0.464 [11]
SVM-LA(β = 0.5) 0.9250 0.649 [11]
Profile(5,7.5) 0.9800 0.794 [10]
SW-PSSM(3.0,0.750,1.50) 0.9820 0.904 [10]
  1. SVM-Ngram, SVM-Pattern, SVM-Motif, SVM-Bprofile and SVM-Top-n-gram refer to the methods based on the five building blocks: N-grams, patterns, motifs, binary profiles and Top-n-grams respectively. The methods with combine suffix refer to the methods combining Top-1-grams and Top-2-grams. The methods with LSA suffix refer to the corresponding methods after latent semantic analysis. Source is the sources of results.